Staff Software Engineer, ML Platform

Summary

Join Stack's ML Training team and contribute to building a reliable, scalable, and user-friendly training framework for Stack's autonomous vehicle (AV) modeling needs. This team is responsible for enhancing the developer experience of ML engineers by creating tools for testing, validation, and model understanding. You will play a key role in model optimization and deployment. The ideal candidate possesses experience with ML platforms, building ML applications, and scalable infrastructure. A deep understanding of design tradeoffs and the ability to collaborate effectively are essential. Prior AV experience is desirable but not required.

Requirements

• Have experience with both ML Platforms and building ML-based applications
• Have experience building scalable, reliable infra at a fast-paced environment
• Have the ability to work across teams
• Have experience building or using ML infra built for a large number of customer teams
• Have a deep understanding of design tradeoffs and ability to articulate those tradeoffs and work with others on getting alignment

Responsibilities

• Provide a reliable, scalable, and easy to use training framework for modeling needs of Stack AV
• Be responsible for the overall developer experience of ML engineers including building tools for testing, validation, and understanding models and the data used to train them
• Be responsible for model optimization and deployment

Preferred Qualifications

• Have experience with building ML models or ML infra in the domains of autonomous vehicles, perception, and decision making
• Have experience with model training, model optimization, or large data processing pipelines
• Know how to push the GPU to its limit from Python to CUDA kernel level
• Have built the inference or training loop for a large model (ideally with LLM flavor)
• Have shipped ML products (NLP, computer vision, recommender systems, etc.) at scale to make business impact
• Know how to build low latency / high throughput batch or stream processing pipelines
• Know how to write (readable) high performance C++
• Have prior AV experience